UNDER FREQUENCY LOAD SHEDDING FOR ENERGY MANAGEMENT USING ANFIS CASE STUDY

Document Sample
UNDER FREQUENCY LOAD SHEDDING FOR ENERGY MANAGEMENT USING ANFIS CASE STUDY Powered By Docstoc
					 International Journal of Electrical Engineering and Technology (IJEET), ISSN 0976 –
INTERNATIONAL JOURNAL OF ELECTRICAL ENGINEERING
 6545(Print), ISSN 0976 – 6553(Online) Volume 4, Issue 2, March – April (2013), © IAEME
                            & TECHNOLOGY (IJEET)

ISSN 0976 – 6545(Print)
ISSN 0976 – 6553(Online)
Volume 4, Issue 2, March – April (2013), pp. 93-104
                                                                              IJEET
© IAEME: www.iaeme.com/ijeet.asp
Journal Impact Factor (2013): 5.5028 (Calculated by GISI)                 ©IAEME
www.jifactor.com




            UNDER FREQUENCY LOAD SHEDDING FOR ENERGY
               MANAGEMENT USING ANFIS/CASE STUDY

                               M.S.Sujatha1, Dr M. Vijay Kumar2
                          1
                         Professor, SVEC, Tirupati-517102, AP, India
                      2
                       Professor, JNTUCEA, Anantapur-515002, AP, India


    ABSTRACT

           Energy management is the major concern for both developing and developed
    countries. Energy sources are scarce and expensive to develop and exploit, hence we
    should confer a procedure to accumulate it by the use of load shedding. If the disturbance
    is large, like large load variations, outage of components (transmission lines, transformers,
    generators, etc.) causes power system blackouts. The conventional method is to solve an
    optimal power flow problem to find out the rescheduling for overload alleviation. But this
    will not give the desired speed of solution. Speed and accuracy of under frequency load
    shedding (UFLS) has a vital role in its effectiveness for preserving system stability and
    reducing energy loss. Initial rate of change of frequency is a fast and potentially useful
    signal to detect the overload when a disturbance accurse. This paper presents a new
    method for solving UFLS problem by using adaptive neuro-fuzzy controller for
    determining the amount of load shed to avoid a cascading outage. The development of new
    and accurate techniques for vulnerability control of power systems can provide tools for
    improving the reliability, continuity of power supply and reducing the energy loss. The
    applicability of ANFIS is tested on a case study at Renigunta 220/132/33 KV sub- station.

    Keywords: Under Frequency Load shedding, Rate of frequency decline, Adaptive Neuro-
    Fuzzy, blackouts.

    1. INTRODUCTION

           When power system is in stable operation at normal frequency, the total mechanical
    power input from the prime movers to the generators is equal to the sum of all running
    loads, plus all real power losses in the system. The frequency conditions of the overall

                                                 93
International Journal of Electrical Engineering and Technology (IJEET), ISSN 0976 –
6545(Print), ISSN 0976 – 6553(Online) Volume 4, Issue 2, March – April (2013), © IAEME

system will directly depend on the amount of active power that the generator prime movers
could deliver to the system. With a small disturbance, the frequency decay rate will be
low and the turbine governor will quickly raise the steam or water to the turbine to
restore the frequency, provided the system has sufficient spinning reserve. However, if the
disturbance is large, like large load variations, outage of components (transmission lines,
transformers, generators, etc.) causes power system blackouts [5]. Recently there was a
major disturbance in Northern Region at 02.33hrs on 30-07-2012. Subsequently, there was
another disturbance at 13.00hrs on 31-07-2012 resulting in collapse of Northern, Eastern and
North-Eastern regional grids [15]. Due to this the frequency may fall to a dangerous value
before the turbine Governor fully operated. The decrease in system frequency, which occurs
very rapidly and will lead to system collapse. The important protection strategy used for
this purpose is a class of protection schemes known as system protection schemes (SPS).
One of the most commonly used type of system protection scheme, generally accepted
after the northeastern blackout of 1965, is Under Frequency Load Shedding (UFLS)
scheme.
        Conventional UFLS system is designed to recover the balance of generation and
consumption following a generator outage or sudden load increase. The loads to be shed
by this system are constant load feeders and are not selected adaptively. In other words,
always the same loads are dropped from the system, regardless of the location of
disturbance. In this method loads are classified in three groups of non-vital, semi vital and
vital loads [11]. However, in recent years, centralized load shedding algorithms to enhance
adaptability of the schemes has been proposed in [4]. Adaptive under-frequency load
shedding based on the magnitude of the disturbance estimation is considered in [9]. S. J.
Huang and C. C. Huang presented Adaptive under-frequency load shedding based on the
time based for remote and isolated power system networks [16].Other adaptive load
shedding methods based on df/dt are presented in [1], [6]-[8]. The work introduces the new
intelligent UFLS using adaptive neuro-fuzzy controller to decide the amount of load to be
shed.

2. CONVENTIONAL LOAD SHEDDING APPROACH

     This section is a review of load shedding techniques that have been devised over a
number of years each having its own set of applications and drawbacks.

2.1 Breaker Interlock Load Shedding

       This is the simplest method of carrying out load shedding. For this scheme, the
circuit breaker interdependencies are arranged to operate based on hardwired trip signals
from an intertie circuit breaker or a generator trip. This method is often used when the
speed of the load shedding is critical. Even though, the execution of this scheme is fast,
breaker interlock load shedding possesses a number of inherent drawbacks.

        •   Load shedding based on worst-case scenario
        •   Only one stage of load shedding
        •   More loads are shed than required
        •   Modifications to the system is costly



                                             94
International Journal of Electrical Engineering and Technology (IJEET), ISSN 0976 –
6545(Print), ISSN 0976 – 6553(Online) Volume 4, Issue 2, March – April (2013), © IAEME

2.2. Under-Frequency Relay Load Shedding

        Guidelines for setting up a frequency load shedding are common to both large and
small systems [12]. The design methodology considers fixed load reduction at fixed system
frequency levels. Upon reaching the frequency set point and expiration of pre-specified
time delay, the frequency relay trips one or more load breakers. This cycle is repeated
until the system frequency is recovered. Drawbacks of this scheme are

            Slow response time.
            Incorrect / Excessive load shedding
            Analysis knowledge is always lost


2.3. Programmable Logic Controller-Based Load Shedding

        A PLC-based load shedding scheme offers many advantages such as the use of an
existing distributed network via the power management system, as well as an automated
means of load relief. However, in such applications monitoring the power system is limited
to a portion of the network with the acquisition of scattered data. This drawback is further
compounded by the implementation of pre-defined load priority tables at the PLC level that
are executed sequentially to curtail blocks of load regardless of the dynamic changes in the
system loading, generation, or operating configuration. The system-wide operating condition
is often missing from the decision-making process resulting in insufficient or excessive load
shedding. In addition, response time (time between the detection of the need for load
shedding and action by the circuit breakers) during transient disturbances is often too long
requiring even more load to be dropped

3. MATHEMATICAL ANALYSIS OF LOAD SHEDDING:

3.1 Active Power and Frequency

       During normal operation of power system, the total mechanical power input to the
system generators is equal to the sum of the connected loads and all real power losses. If, for
any reason, the balance of generation and loads is disrupted, operating frequency of the
system would change according to equation (1).

     df
2H      = Pm − Pe                                                                         (1)
     dt

Where H is the inertia constant, f is the generator frequency in pu, Pm is the generator
mechanical power in pu and Pe is the generator electrical power in pu.
       In general, power system loads are dependent on both voltage and frequency. To
compute exact amount of system load, this dependence should be considered. Power system
voltage is a local parameter and its value is not usually known precisely for each remote
system bus. As such, the effect of voltage in load modeling has been ignored in several
researches. Despite of voltage, system frequency is a general parameter and is almost same
at all points of a power system. The frequency dependent characteristic of composite load is

                                              95
International Journal of Electrical Engineering and Technology (IJEET), ISSN 0976 –
6545(Print), ISSN 0976 – 6553(Online) Volume 4, Issue 2, March – April (2013), © IAEME

expressed in equation (2).

Pe = PL + d ∆f                                                                            (2)
 Where, d is load reduction factor and PL is the load power in Pu.

3.2 Final Frequency

       According to equations (1) and (2) if, a power system encounters with an overload
case, and then frequency of the system begins to decline. At the same time, system load
which is dependent on frequency by factor d is reduced as well. If no other action is
performed, eventually system settles at a final frequency at which the amount of
generation reduction would be equal to reduction of system loads. The final frequency is
given by the equation (3).

                        
                                l0       
                                          
            f ∞ = f 0 − 1 −              
                            (1 + l 0 ) d                                                (3)
                                         
Where, f0 is the initial normal frequency, f∞ is the final frequency and l0 is the amount of
overload in pu.

3.3 Minimum Allowable Frequency

All of the power system apparatus are made and designed for nominal frequency. Power
plant auxiliary services are more demanding in terms of minimum allowable frequency.
Steam turbine is the most sensitive equipment against frequency drops. Continuous
operation of steam turbines should be restricted to frequency above 47.5Hz. Frequency falls
below 47 Hz must be avoided. In fact, every commercial turbine can sustain up to 10
contingencies at 47 Hz just for one second without being jeopardize [13].

 3.4. Conventional under frequency load shedding scheme

       For a system overload, the amount of load which must be shed is determined based
 on the minimum allowable frequency and the amount of overload. The minimum
 allowable frequency could be considered as 47 Hz instead of 47.5Hz in load shedding
 schemes design, if the system dispatch center can increase the generation by governors
 quickly. Total amount of the load which must be shed to cover the maximum anticipated
 overload, is obtained from equation (4) is employed in [10].

      L            f 
          − d 1 − 
LD =
     1+ L      50                                                                        (4)
                 f 
      1 − d 1 − 
             50 
   Where, LD is the total load which must be shed, L is the rate of overload per unit
produced in the system, f is permissible settling frequency and the d coefficient is the load
reduction coefficient which is related to the type of disturbance.


                                              96
International Journal of Electrical Engineering and Technology (IJEET), ISSN 0976 –
6545(Print), ISSN 0976 – 6553(Online) Volume 4, Issue 2, March – April (2013), © IAEME

4. PROPOSED SYSTEM
       Proposed system discusses the neural network, fuzzy logic and adaptive neuro fuzzy
methods of load shedding.

4.1. Neural network
         Fast and optimal adaptive load shedding method for power system is using, Artificial Neural
Networks [2], [14]. The first step in the design of a NN is to determine an architecture that will yield
good results. The idea is to use the simplest architecture while maximizing performance. Usually,
NN architecture is determined based on subjective assessment on the part of the engineer. In
order to achieve appropriate neural network training lots of neural networks have been trained
and concluded that the architecture of 2 hidden layers, the first with 14 nodes and the second with
10 nodes, was best suited for this application.
         Levenberg Marquardt Back Propagation technique [3] is used to train the NN. This consists
of the same routine as typical back propagation with the exception that instead of one learning rate
for all the NN nodes, a learning rate was assigned to each of the nodes in order to speed up
convergence. The activation function used within this work is a hyperbolic tangent function and the
inputs were normalized to have a mean of zero and a variance of one.

4.2 Fuzzy Logic
        The generation and load data are given as inputs to the simulation system shown in Figure 1,
it calculates the rate of change of frequency at that moment and sends that data to the fuzzy
controller. The fuzzy controller decides the amount of load shed required to maintain the system to
operate in stable manner as shown in Table.3 [17].

4.3. Neuro-Fuzzy
        Neuro-fuzzy system is a combination of neural network and fuzzy logic in which it
combines the learning and adapting abilities of neural networks with the fuzzy interpretation of fuzzy
logic system.
        A typical ANFIS structure with five layers and two inputs, each with two membership
functions and Rule viewer in ANFIS approach is shown in Figures 2 and 3 respectively. The ANFIS
has the constraint that it only supports the Sugeno-type systems of 1st or 0th order. The system can
only be designed as a single output system and the system must be of unity weights for each rule.




                      Fig.1: Simulation diagram of Fuzzy Controlled Load Shedding


                                                   97
International Journal of Electrical Engineering and Technology (IJEET), ISSN 0976 –
6545(Print), ISSN 0976 – 6553(Online) Volume 4, Issue 2, March – April (2013), © IAEME




         Fig.2: ANFIS Structure with 1input 1 Output, and 1 Membership Function for each
                                              input

         Fuzzy logic system needs rules to be defined first, but one may have no knowledge
about a power system for the formation of rules. Therefore, automated parameters tuned by
a neural network embedded inside a fuzzy system can replace the need for prior knowledge
about a power system. To implement the proposed load shedding scheme for vulnerability
control of power systems, firstly, base case simulation is carried out on a power system so as
to analyze the system behavior at the base case condition. The next step is to analyze the
system behavior when subjected to credible system contingencies such as line outage (LO),
generator outage (GO), load increase (LI) and disconnection of loads (DL). The selection of
inputs to the controller is an important design consideration and therefore for power system
vulnerability control, the rate of change of frequency is selected as input variable for the
load shedding controller. The overall output of the ANFIS is the estimated amount of load
shed. Here the training data is given which was collected from the Electricity Board and
error is observed.




                           Fig.3: Rule viewer in ANFIS Approach

5. SIMULATION RESULTS

        Figure 4 shows the training error of neuro fuzzy controller. Figures 5a, 5b and 5c
show the variation of frequency during normal condition, load is greater than generation and
after restoration using proposed technique respectively.. Table 1 shows the inputs of the
fuzzy controller (Generation and Load) by that it calculates the rate of change of frequency

                                              98
International Journal of Electrical Engineering and Technology (IJEET), ISSN 0976 –
6545(Print), ISSN 0976 – 6553(Online) Volume 4, Issue 2, March – April (2013), © IAEME

(df/dt) and the output obtained is the amount of load shed required to maintain the system to
operate in a stable manner. Table 2 shows comparison of actual load shed using conventional
approach with fuzzy and neuro-fuzzy.




                        Fig.4: Training Error of Neuro-fuzzy controller




                           Fig.5: Variation of frequency under different
                                   operating conditions

   Table 2: Comparison of load shed values of AP grid with neural, Fuzzy and Neuro fuzzy
                                         systems

                                                                             Neuro
                       Actual value       Neural shed        Fuzzy
             df/dt                                                         fuzzy shed
                      shed by AP grid       value          shed value
                                                                              value
              -.3           120                123             119             119
              -.9           360                357             344             358
             -1.4           560                554             536             557
             -1.8           720                717             695             717
             -2.6          1040               1042            1000            1039




                                              99
International Journal of Electrical Engineering and Technology (IJEET), ISSN 0976 –
6545(Print), ISSN 0976 – 6553(Online) Volume 4, Issue 2, March – April (2013), © IAEME

6. CASE STUDY

       In the case study 220/132/33KV substation, Renigunta, AP is considered. It
consists of 13 numbers of 132 KV, 6     numbers of 20KV and 7 numbers of 3KV
feeders.

        Table 3 & Table 4 shows the import and export details of Renigunta 220/132/33
KV substation respectively. By observing here, it is clear that the generation is lower
than load, so load shedding must require to protect the grid from collapse and reduce
the drastically decrease in frequency. By using these Import and Export details, neural
network is trained and the amount of load shed required for maintaining the grid in stable
was observed. In this work the generation and load data are given as inputs to the
simulation system, it calculates the rate of change of frequency at that moment and sends
that data to the fuzzy controller. The fuzzy controller decides the amount of load shed
required to maintain the system to operate in stable manner. Table 5 shows that, the
amount of load to be shed by fuzzy and neuro – fuzzy using rate of change of frequency
as the input to the controller.

       From the Table 6, it is observed that the required amount of load to be shed is less
than the existing method of load shedding value. The amount of load shed by neural,
fuzzy and neuro – fuzzy approach is very low when compared to existing method of load
shed value. By this it is clear that, using proposed method energy is saved, frequency is
restored and power can be supplied for more number of consumers.

              Table 3: Import details of Renigunta 220/132/33 KV sub-station

                              Feeder name          Input power


                              Manubolu-1             55MW


                              Manubolu-2             55MW

                                CK Palli             135MW

                                 Kodur               70MW


                                  Total
                                                     315MW




                                            100
International Journal of Electrical Engineering and Technology (IJEET), ISSN 0976 –
6545(Print), ISSN 0976 – 6553(Online) Volume 4, Issue 2, March – April (2013), © IAEME

             Table 4: Export details of Renigunta 220/132/33 KV sub-station

                      Feeder name                 Output value

                      M.Mngalam-1                   72 MW

                      M.Mngalam-2                   72 MW

                        Railway-1                       0 MW

                        Railway-1                       6 MW

                        Grindwel                    10 MW

                        Amaraja                      16MW

                        Puttur-1                     12MW

                        Puttur-2                     12MW

                        Tirupathi                    50MW

                       Chandragiri                   17MW

                            Lv-1                   17.6MW

                            Lv-2                   12.4MW

                                                    324MW
                            Total



            Table 5: Comparisons of Actual, Neural, Fuzzy, and Neuro –
                              Fuzzy output values

                    df/dt       Actual           Fuzzy            Neuro-
                              difference         output           fuzzy
                                                                 output
                    -0.2             8              7              7.8

                    -0.4            16             14             15.6

                    -0.5            20             17             19.5

                    -0.7            28             24             27.5




                                           101
International Journal of Electrical Engineering and Technology (IJEET), ISSN 0976 –
6545(Print), ISSN 0976 – 6553(Online) Volume 4, Issue 2, March – April (2013), © IAEME

                      Table 6: Comparing real time and project values

            Required            Actual         Neural      Fuzzy          Neuro-
         amount of Load         Load           output      output       fuzzy output
           shed(MW)              shed          (MW)        (MW)            (MW)
                               value(M
                12                w)
                                  28              12        10              11

                20                28              19        17              19

                27                36              26        23              26

                35                40              33        30              33



7. CONCLUSIONS

        Due to the shortage of electricity, load shedding is extremely common in
India. To overcome drastically decreasing frequency of the system, usually, load
shedding is performed. Automatic load shedding is required to anticipate and relieve the
overloaded equipment before there is loss of generation, line tripping, equipment damage, or
a chaotic random shutdown of the system.
        The present work deals with under frequency load shedding. It shows greater
importance in maintaining the frequency in a stable and constant manner. Stability is
mainly based on recovering, rate of change of frequency to normal value. To avoid this
drastically decrease of rate of change of frequency an ANFIS controller is designed to find
the amount of load to be shed at demand side in order to bring back the frequency to a
normal value .
        The existing methods such as Breaker inter locking system, Under Frequency Relay
Load Shedding, PLC based controllers are used for the load shedding. These methods are
time consuming and also shed excess amount of load. Compared to other under frequency
load shedding control methods, the ANFIS method/approach shows the advantages like
lower value of load shedding, shorter response of time, saving more amounts of energy
and benefiting more consumers. The applicability of ANFIS is tested on a case study at
Renigunta 220/132/33 KV sub- station.

REFERENCES

[1]   M. Moazzami and A. khodabakhshian, “A New Optimal Adaptive under Frequency
      Load Shedding Using Artificial Neural Networks,” Proceedings of ICEE 2010, May
      11-13.
[2]   Pukar Mahat, Zhe Chen and Birgitte Bak-Jensen, “Under frequency Load
      Shedding for an Islanded Distribution System with Distributed Generators,” IEEE
      transactions on power delivery, vol.25, no.2, April. 2010.



                                            102
International Journal of Electrical Engineering and Technology (IJEET), ISSN 0976 –
6545(Print), ISSN 0976 – 6553(Online) Volume 4, Issue 2, March – April (2013), © IAEME

[3]     C.-T. Hsu, M.-S. Kang and C.-S. Chen “Design of adaptive load shedding artificial
       neural networks “IEE Proc.-Gener. Transm. Distrib., Vol. 152, No. 3, May. 2005
[4]     M.Sanaye Pasand, H..Seyedi, “New Centralized Adaptive under Frequency Load
        Shedding Algorithms,” Large Power Engineering System conference, 10-12 Oct. 2007,
        pp. 44-48.
[5]     M. Sanaye Pasand, M. R. Dadashzadeh, “Iran national grid blackout, power system
        protection point of view,” Developments in Power System Protection, 8‘Th IEEE
        International Conf. Vol. 1, April.2004, pp 20-23.
[6]     P. M. Anderson and M..Mirheydar, “An adaptive method for setting under frequency
         load shedding relays,” IEEE Trans. Power Syst., vol.7, no. 2, May.1992, pp. 647–655.
[7]     S. J. Huang and C. C. Huang, “An adaptive load shedding method with time-based
        design for isolated power systems,” Int. J. Elect. Power Energy Syst., vol. 22, Jan. 2000,
        pp. 51–58.
[8]     D. Prasetijo, W. R. Lachs, and D. Sutanto, “A new load shedding scheme for limiting
         under frequency,” IEEE Trans. Power Syst., vol. 9, no. 3, Aug. 1994, pp. 1371–1378.
[9]     V. V. Terzija, “Adaptive under-frequency load shedding based on the magnitude
         of the disturbance estimation,” IEEE Trans. Power Syst., vol. 21, no. 3, Aug.2006, pp.
         1260–1266.
[10]    M. R. Dadashzadeh, M. Sanaye-Passand, “Simulation and Investigation of Load
         Shedding Algorithms for a Real Network Using Dynamic Modeling”, 39th International
         Universities Power Engineering conf. 2, Sept. 2004, pp. 1111-15.
[11]    Z. Ding, D. Cartes and S. Sirvastava, “New load shedding scheme for islanded
        power systems,” 2006 IEEE/SMC International Conference on System of Systems
        Engineering Los Angeles, CA, USA - April.2006.
[12]    J. R. Jones and W.D. Kirkland, "Computer algorithm for selection of frequency relays
         for load shedding," IEEE Computer Application in Power, Jan. 1988, pp. 21-25.
[13]    M. NagpaI, A. Moshref, Z. Shukn, .I. T. Vijayan, D. Chan and C. Henvile "Dynamic
         simulation optimize application of an under frequency load shedding scheme," Western
         Protective Relay Conference, Spokane, Washington, October 2 1-23, 1997.
[14]    N.ManojKumar, M.S.Sujatha, T.Devaraj, N.M.G.Kumar “Artificial neural network
         approach for under frequency load shedding “International journal of scientific and
         engineering research, vol 3, issue 7, July .2012.
[15]   “Report enquiry committee on grid Disturbance in Northern region on 30th July 2012
         and in and North-Eastern on 31th July 2012” http//www.powermin.nic.in, 16 August
         2012, Retrieved 28 August 2012
[16]   S. J. Huang and C. C. Huang, “An adaptive load shedding method with time-based
       design for isolated power systems,” Int. J. Elect. Power Energy Syst., vol. 22,
       pp. 51–58, Jan. 2000.
[17]   M.S.Sujatha, Manoj Kumar.N and Dr M. Vijay Kumar, “Under Frequency Load Shedding
       for Reduction of Energy Losses by Adaptive Neuro Fuzzy Technique”, International
       Journal of Computer Engineering and Technology, Volume 3, Issue 2, 2012, pp. 12 - 18,
       ISSN Print: 0976 – 6367, ISSN Online: 0976 – 6375.
[18]   Mukesh Kumar Kirar and Ganga Agnihotri, “Transient Stability Enhancement By Ann
        Based Adaptive Load Shedding” International Journal of Electrical Engineering &
        Technology (IJEET), Volume 3, Issue 3, 2012, pp. 200 - 210, ISSN Print : 0976-6545,
        ISSN Online: 0976-6553



                                               103
International Journal of Electrical Engineering and Technology (IJEET), ISSN 0976 –
6545(Print), ISSN 0976 – 6553(Online) Volume 4, Issue 2, March – April (2013), © IAEME

AUTHORS BIBLIOGRAPHY

             M.S.Sujatha is currently pursuing her PhD degree in Electrical and
             Electronics Engineering department at university of JNTUCA, Anantapur,
             A.P, India. She received her B.E degree from Mysore University and
             M.Tech degree from JNTUCA, Anantapur. Her research interest includes
             wireless technologies for power system applications; protection, monitoring
             and control and reduction in Energy Losses.


               Dr. M. Vijay Kumar Graduated from NBKR Institute of Science and
               Technology, Vidyanagar, A.P, India in 1988.He obtained M.Tech degree from
               Regional Engineering College, Warangal, India in 1990. He received
               Doctoral degree from JNT University, Hyderabad, India in 2000. He has
guided 15 PhD’s. Currently, he is Professor in Electrical and Electronics Engineering
Department at JNTUA, Anantapur A.P, India. His areas of research interest are power
system protection, monitoring, control, power electronic and drives and power quality
issues. He is the recipient of The Pandit Madan Mohan Malaviya Memorial Prize.




                                          104

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:0
posted:4/5/2013
language:
pages:12